基于表面肌电非负矩阵分解算法的鸽子机器人后肢运动控制策略研究  

Research on pigeon robot hind limb motion control strategy based on surface EMG nonnegative matrix decomposition algorithm

作　　者：姚文桢 张淋皓 常铭 张宇 槐瑞托 杨俊卿 常辉 汪慧 YAO Wenzhen;ZHANG Linhao;CHANG Ming;ZHANG Yu;HUAI Ruituo;YANG Junqing;CHANG Hui;WANG Hui(College of Electrical Engineeringand Automation,Shandong University of Science and Technology,Qingdao 266400,China)

机构地区：[1]山东科技大学电气与自动化工程学院,山东青岛266400

出　　处：《现代电子技术》2024年第12期129-137,共9页Modern Electronics Technique

基　　金：国家自然科学基金项目(61903230);山东省自然科学基金项目(ZR202103010716);山东省自然科学基金项目(AR2020MF098)。

摘　　要：为了深入研究鸽子机器人运动控制神经机制,分析了自由运动状态和电刺激背侧丘间核状态下,鸽子后肢伸缩腿动作的肌肉协同性及相似性。采集后肢表面肌电信号进行预处理,通过非负矩阵分解算法分别将自由运动状态和电刺激背侧丘间核运动状态下的肌肉协同分解为肌肉贡献度矩阵和肌肉激活系数矩阵;然后使用皮尔逊相关系数计算相似性。结果表明,与自由运动状态相比,电刺激背侧丘间核时分解出的肌肉贡献度值更高,肌肉激活时间更早,激活时间更长,肌肉贡献度矩阵及激活系数矩阵的相似值更大。这说明在电刺激状态下,较多的肌纤维更早产生兴奋,肌肉更快、更强收缩,神经系统对协同肌的控制作用更强,使运动行为更加协调有序,从而能快速有效地产生强烈的行为反应以应对刺激。In order to further research the neural mechanism of motion control of pigeon robot,the muscular synergism and similarity of the retractor leg movement of pigeon hind limb were analyzed under the condition of free movement and electric stimulation of dorsal intercollicular nucleus.The surface EMG signals of the hind limbs were collected for pre-processing,and the non-negative matrix factorization algorithm is used to decompose the muscle synergy under free motion and electrical stimulation of the dorsal colliculus into muscle contribution matrix and muscle activation coefficient matrix,respectively.Pearson correlation coefficient is used to calculate the similarity.The results show that in comparison with the free motion state,the muscle contribution value decomposed by electrical stimulation of the dorsal interchalamic nucleus was higher,the muscle activation time was earlier,the activation time was longer,and the similarity values of the muscle contribution matrix and activation coefficient matrix were larger.It indicates that under electrical stimulation,more muscle fibers generate excitement earlier,muscles contract faster and stronger,and the nervous system has a stronger control over the synergistic muscles,making motor behavior more coordinated and orderly,thus enabling strong behavioral responses to respond to stimuli quickly and effectively.

关 键 词：鸽子机器人 后肢运动控制 肌肉协同 非负矩阵算法 背侧丘间核 相似性分析 

分 类 号：TN923.3-34[电子电信—通信与信息系统] TP242[电子电信—信息与通信工程]